Delirium is a common and potentially devastating problem for hospitalized elderly patients, with occurrence rates from 25-60% and hospital mortality rates from 25-33%. Previous studies have estimated that if the length of stay of each hospitalized elderly delirious patient could be reduced by just one day, the savings to Medicare would be over $1-2 billion per year. The present trial is proposed to address this compelling but neglected clinical problem. The primary aim is to determine whether a multicomponent targeted intervention (MTI) strategy reduces the risk of development of delirium as compared with usual care (UC). A related aim is to estimate the effect of the MTI strategy on delirium duration, recurrence, and severity, nursing home placement, and death. Secondary aims are to determine adherence to and cost-effectiveness of the MTI strategy. The proposed study is a controlled clinical trial with matched stratified block design involving 560 subjects, who are greater than or equal to age 70 years and at moderate to high risk for delirium at admission (based on a previously developed risk stratification system), enrolled from one MTI and three UC medical units. A strategy of prospective matching will be used to ensure comparable patients on the MTI and UC units; the extra UC units will provide sufficient control patients to facilitate the matching process. The MTI strategy, applied from admission until discharge, will be targeted towards specific delirium risk factors (i.e., cognitive impairment, vision or hearing impairment, dehydration, psychoactive medications, sleep deprivation, and immobilization) and will include direct patient, nursing-related, and environmental interventions. Delirium will be assessed daily from admission until discharge and at one month follow-up using the standardized, validated Confusion Assessment Method; and will be defined according to the presence of: l) acute onset and fluctuating course; 2) inattention; and either 3) disorganized thinking; or 4) altered level of consciousness. Outcomes will be assessed by research staff not involved in the intervention, and blinded to the subjects' group assignment and to the interventional nature of the study. The study was designed to ensure adequate power (greater than or equal to 80%) to detect a reduction of incident delirium from 25% in UC to 16.75% in MTI patients, a 33% absolute reduction. Unique advantages of this study include the testing of practical, real- world solutions for prevention of delirium; the daily, systematic assessment of delirium, using a validated instrument; the targeting of patients at moderate to high risk for delirium, maximizing study efficiency and beneficial impact; and the multidisciplinary expertise of the research team. This trial is highly significant for the promise it holds for reducing delirium and its associated morbidity and mortality, and for its potential health benefits to the geriatric population at large.